随着对无线传感器网络的广泛研究与应用,用户对传感器节点的安全性要求日益提高。由于传统基于密码学的信息安全技术并不能够完美地解决传感器节点面临的复杂安全威胁,信誉系统已经被引入到无线传感器网络中,对节点安全情况进行周期性评估,并分配相应的信誉值。由于信誉系统很难分辨一些节点的某些行为是否处于正常区间如监测数据是否准确,进而导致某些传感器节点能够躲避信誉系统的监测,对用户决策产生不良影响。本文提出了一种结合信誉系统和噪声点检测技术的无线传感器网络节点安全模型。一方面,网络中的信誉系统模块为噪声点检测模块提供数据支撑,以便高效检测到噪声点数据;另一方面,噪声点检测模块对信誉系统进行反馈,加速节点信誉值的收敛,提高系统效率。一系列的仿真表明,相比于传统信誉系统模型,改进后的节点安全模型能够同时检测到网络攻击和数据攻击,同时该模型具有更高的收敛速度。 With the extensive research and application of wireless sensor networks, the security requirements of the sensor nodes are increasing day by day. Because the traditional cryptography-based information security technology can not perfectly solve the complex security threats faced by sensor nodes, the reputation system has been introduced into the WSN. The node security situation is periodically evaluated and the corresponding credit value is assigned. It is difficult for credit system to distinguish whether certain behaviors of some nodes are in normal intervals, such as whether the monitoring data is accurate or not, so that some sensor nodes can evade the monitoring of reputation system and have an adverse effect on user decision-making. This paper presents a wireless sensor network node security model combined with reputation system and noise point detection technology. On the one hand, the credit system module in the network provides the data support for the noise point detection module so as to detect the noise point data efficiently; on the other hand, the noise point detection module feeds back the credit system to accelerate the convergence of the credit value of the node and improve the system efficiency . A series of simulations show that compared with the traditional reputation system model, the improved node security model can detect both network attacks and data attacks at the same time, and the model has higher convergence speed.